import pandas as pd
from sklearn.preprocessing import MultiLabelBinarizer, OneHotEncoder, LabelEncoder
import matplotlib.pyplot as plt
import seaborn as sns
import plotly.express as px
from scipy.stats import iqr
import scipy.stats as stats
import numpy as np
import plotly.graph_objects as go
import re
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
from sklearn.metrics import mean_squared_error, r2_score
from sklearn.linear_model import Ridge
from sklearn.ensemble import RandomForestRegressor

data = pd.read_excel('/content/final_dataset.xlsx')

data.head()

data.describe()

data.isnull().sum()

Player                                    0
Club                                      0
Age                                       0
Position                                  0
Nation                                    0
                                       ... 
Own Goals (17/18)                       921
Total Loose Balls Recovered (17/18)     919
Aerial Duel Won (17/18)                 919
Aerial Duel Lost (17/18)                919
% Aerial Duels Won (17/18)             1023
Length: 548, dtype: int64

duplicate_mask = data.duplicated(subset=['Player', 'Club', 'Age'], keep='first' )

non_unique_values = data[duplicate_mask]

non_unique_values

data = data.drop(non_unique_values.index)

data['G + A (20/21)'] = data['Gls (20/21)'] + data['Ast (20/21)']

#confirmed they different
data[['G + A (20/21)', 'Goal Creating Actions (20/21)']]

# Handle missing values

def split_squad(squads):
    if isinstance(squads, str):
        return squads.split(', ')
    elif isinstance(squads, list):
        return squads
    elif isinstance(squads, (int, float)):
        return [str(squads)]
    else:
        return [str(squads)]

# Apply for each "Squad" column
squad_columns = [col for col in data.columns if 'Squad' in col]

mlb_filled = MultiLabelBinarizer()
mlb_dropped = MultiLabelBinarizer()

# Initialize the encoders
mlb = MultiLabelBinarizer()
le_nation = LabelEncoder()
le_position = LabelEncoder()
le_league = LabelEncoder()


def encode_dataframe(df):
    encodings = {
        'Nation': {},
        'Position': {},
        'League': {}
    }

    for col in [c for c in df.columns if 'Squad' in c]:
        df[col] = df[col].apply(split_squad)
        squad_encoded = pd.DataFrame(mlb.fit_transform(df[col]), columns=[f"{col}_{squad}" for squad in mlb.classes_], index=df.index)
        df = df.join(squad_encoded).drop(col, axis=1)

    if 'Nation' in df.columns:
        df['Nation'] = le_nation.fit_transform(df['Nation']) + 1
        encodings['Nation'] = {index + 1: label for index, label in enumerate(le_nation.classes_)}

    if 'Position' in df.columns:
        df['Position'] = le_position.fit_transform(df['Position']) + 1
        encodings['Position'] = {index + 1: label for index, label in enumerate(le_position.classes_)}

    if 'League' in df.columns:
        df['League'] = le_league.fit_transform(df['League']) + 1
        encodings['League'] = {index + 1: label for index, label in enumerate(le_league.classes_)}

    return df, encodings

encoded_data, encodings = encode_dataframe(data)

data_filled = encoded_data.fillna(0)
data_dropped = encoded_data.dropna()

data_filled.head()

data_dropped.head()

data_for_corr_filled = data_filled.drop(columns = ["Player", "Club"])
data_for_corr_filled.head()

def get_top_n_corr_for_feature(corr_matrix, target_feature, n):
    target_corr = corr_matrix[target_feature]

    # Get the top n features with the highest absolute correlation values (excluding the target feature itself)
    top_n_features = target_corr.abs().sort_values(ascending=False).drop(target_feature).head(n).index

    # Include the target feature in the top features
    top_n_features = list(top_n_features)
    top_n_features.append(target_feature)

    # Create a reduced correlation matrix with the target feature and the top 20 features
    reduced_corr_matrix = corr_matrix.loc[top_n_features, top_n_features]

    return reduced_corr_matrix

def plot_reduced_corr_matrix(reduced_matrix, target_feature, n):
    plt.figure(figsize=(18, 16))
    sns.heatmap(reduced_matrix, annot=True, cmap='coolwarm', vmin=-1, vmax=1, annot_kws={"size": 8})
    plt.xticks(rotation=90)
    plt.yticks(rotation=0)
    plt.title(f'Correlation Matrix of Top {n} Features Correlated with {target_feature}')
    plt.show()

def remove_fail_rows(df):
  # Extract the rows where the scraping failed for contract length
  mask = df.isin(['fail'])
  failed_rows = df[mask.any(axis=1)]
  len(failed_rows)
  rows_to_drop = df[mask.any(axis=1)].index
  return df.drop(rows_to_drop)

target_features = ['Value', 'Position', 'Gls (20/21)', 'Yellow Cards (20/21)']
corr_matrix = remove_fail_rows(data_for_corr_filled).corr()
for target_feat in target_features:
 red_corr_matrix = get_top_n_corr_for_feature(corr_matrix, target_feat, 40)
 plot_reduced_corr_matrix(red_corr_matrix, target_feat, 40)

data_for_corr_dropped = data_dropped.drop(columns = ["Player", "Club"])
data_for_corr_dropped.head()

corr_matrix = remove_fail_rows(data_for_corr_dropped).corr()
for target_feat in target_features:
 red_corr_matrix = get_top_n_corr_for_feature(corr_matrix, target_feat, 40)
 plot_reduced_corr_matrix(red_corr_matrix, target_feat, 40)

plt.figure(figsize=(10, 6))
sns.countplot(x='Position', data=data)
plt.title('Distribution of Players by Position')
plt.xlabel('Position')
plt.ylabel('Count')
plt.xticks(rotation=45)
plt.show()

# Count plot of players by position for each league
plt.figure(figsize=(14, 8))
sns.countplot(x='Position', hue='League', data=data)
plt.title('Breakdown of Player Positions by League')
plt.xlabel('Position')
plt.ylabel('Count')
plt.legend(title='League')
plt.xticks(rotation=45)
plt.show()

# Violin plot of age by position
plt.figure(figsize=(14, 8))
sns.violinplot(x='Position', y='Age', hue='Position', data=data)
plt.title('Distribution of Player Age by Position')
plt.xlabel('Position')
plt.ylabel('Age')
plt.xticks(rotation=45)
plt.show()

def identify_outliers(df, x_col, y_col):
    df['Outlier'] = False  # Add a new column to flag outliers
    for pos in df[x_col].unique():
        subset = df[df[x_col] == pos]
        q1 = subset[y_col].quantile(0.25)
        q3 = subset[y_col].quantile(0.75)
        iqr = q3 - q1
        fence_low = q1 - 1.5 * iqr
        fence_high = q3 + 1.5 * iqr
        outlier_condition = (subset[y_col] < fence_low) | (subset[y_col] > fence_high)
        df.loc[outlier_condition & (df[x_col] == pos), 'Outlier'] = True
    return df

data = identify_outliers(data, 'Position', 'Gls (20/21)')

fig = px.box(data, x='Position', y='Gls (20/21)', color='Outlier', color_discrete_map={True: 'red', False: 'blue'},
             hover_data=['Player'], points='all')
fig.update_layout(
    title='Goals (20/21) by Position',
    xaxis_title='Position',
    yaxis_title='Gls (20/21)'
)

fig.show()

# Calculate summary statistics for outliers
def outlier_summary(df, category_col, y_col):
    summary = []

    # Group by the category and apply calculations only on flagged outliers
    grouped = df[df['Outlier']].groupby(category_col)

    for name, group in grouped:
        summary.append({
            'Category': name,
            'Num_Outliers': group.shape[0],
            'Mean_Outlier_Value': group[y_col].mean()
        })

    if not summary:  # If no outliers are found, create a DataFrame with the correct columns but empty
        return pd.DataFrame(columns=['Category', 'Num_Outliers', 'Mean_Outlier_Value'])
    else:
        return pd.DataFrame(summary)

# Get outlier summary
outlier_summary_df = outlier_summary(data, 'Position', 'Gls (20/21)')
print(outlier_summary_df)

# Plot summary statistics
plt.figure(figsize=(10, 6))
sns.barplot(x='Category', y='Num_Outliers', data=outlier_summary_df)
plt.title('Number of Outliers by Position')
plt.xlabel('Position')
plt.ylabel('Number of Outliers')
plt.show()

     Category  Num_Outliers  Mean_Outlier_Value
0    Defender            60            3.766667
1  Goalkeeper             2            1.000000
2      attack            21           23.857143
3    midfield            28            9.857143

data = identify_outliers(data, 'Position', 'Value')
fig = px.box(data, x='Position', y='Value', color='Outlier', color_discrete_map={True: 'red', False: 'blue'},
             hover_data=['Player'], points='all')
fig.update_layout(
    title='Value by Position',
    xaxis_title='Position',
    yaxis_title='Value'
)

fig.show()

outlier_summary_df = outlier_summary(data, 'Position', 'Value')
print(outlier_summary_df)

# Plot summary statistics
plt.figure(figsize=(10, 6))
sns.barplot(x='Category', y='Num_Outliers', data=outlier_summary_df)
plt.title('Number of Outliers by Position')
plt.xlabel('Position')
plt.ylabel('Number of Outliers')
plt.show()

     Category  Num_Outliers  Mean_Outlier_Value
0    Defender            72        3.761250e+07
1  Goalkeeper            19        3.145263e+07
2      attack            45        6.160000e+07
3    midfield            40        5.539500e+07

data = identify_outliers(data, 'Position', 'Ast (20/21)')
fig = px.box(data, x='Position', y='Ast (20/21)', color='Outlier', color_discrete_map={True: 'red', False: 'blue'},
             hover_data=['Player'], points='all')
fig.update_layout(
    title='Assists by Position',
    xaxis_title='Position',
    yaxis_title='Ast (20/21)'
)

fig.show()

outlier_summary_df = outlier_summary(data, 'Position', 'Ast (20/21)')
print(outlier_summary_df)

# Plot summary statistics
plt.figure(figsize=(10, 6))
sns.barplot(x='Category', y='Num_Outliers', data=outlier_summary_df)
plt.title('Number of Outliers by Position')
plt.xlabel('Position')
plt.ylabel('Number of Outliers')
plt.show()

     Category  Num_Outliers  Mean_Outlier_Value
0    Defender            77            4.545455
1  Goalkeeper             7            1.000000
2      attack            27           10.370370
3    midfield            22            9.954545

data = identify_outliers(data, 'Position', 'Non-Penalty Goals (20/21)')
fig = px.box(data, x='Position', y='Non-Penalty Goals (20/21)', color='Outlier', color_discrete_map={True: 'red', False: 'blue'},
             hover_data=['Player'], points='all')
fig.update_layout(
    title='Non-Pen Goals by Position',
    xaxis_title='Position',
    yaxis_title='Non-Penalty Goals (20/21)'
)

fig.show()

outlier_summary_df = outlier_summary(data, 'Position', 'Non-Penalty Goals (20/21)')
print(outlier_summary_df)

# Plot summary statistics
plt.figure(figsize=(10, 6))
sns.barplot(x='Category', y='Num_Outliers', data=outlier_summary_df)
plt.title('Number of Outliers by Position')
plt.xlabel('Position')
plt.ylabel('Number of Outliers')
plt.show()

     Category  Num_Outliers  Mean_Outlier_Value
0    Defender            58            3.775862
1  Goalkeeper             1            1.000000
2      attack            15           21.200000
3    midfield            19            9.157895

# Density plot of player values
plt.figure(figsize=(10, 6))
sns.kdeplot(data['Value'], shade=True)
plt.title('Density Plot of Player Values')
plt.xlabel('Value')
plt.ylabel('Density')
plt.show()

<ipython-input-29-a020f4c50887>:3: FutureWarning:


`shade` is now deprecated in favor of `fill`; setting `fill=True`.
This will become an error in seaborn v0.14.0; please update your code.

# Pair plot of selected features
selected_features = ['Value', 'Age', 'Gls (20/21)', 'Ast (20/21)']
sns.pairplot(data[selected_features])
plt.suptitle('Pair Plot of Selected Features', y=1.02)
plt.show()

# Scatter plot of value vs. goals
plt.figure(figsize=(10, 6))
sns.scatterplot(x='G + A (20/21)', y='Value', data=data)
plt.title('Value vs. Goals (20/21)')
plt.xlabel('Goals (20/21)')
plt.ylabel('Value')
plt.show()

value_to_find = 'fail'

mask = data.isin([value_to_find])

# Extract the rows where the scraping failed for contract length
failed_rows = data[mask.any(axis=1)]

len(failed_rows)
rows_to_drop = data[mask.any(axis=1)].index
data_for_this_plot = data.drop(rows_to_drop)

# Scatter plot of contract years left vs. value
plt.figure(figsize=(10, 6))
sns.scatterplot(x='Contract Years Left', y='Value', data=data_for_this_plot)
plt.title('Contract Years Left vs. Value')
plt.xlabel('Contract Years Left')
plt.ylabel('Value')
plt.show()

top_nations = data['Nation'].value_counts()

nation_counts = data['Nation'].value_counts()
chunk_size = 23
total_nations = len(nation_counts)
num_chunks = (total_nations + chunk_size - 1) // chunk_size
for i in range(num_chunks):
    start_idx = i * chunk_size
    end_idx = start_idx + chunk_size
    top_nations = nation_counts.index[start_idx:end_idx]

    plt.figure(figsize=(14, 8))
    sns.boxplot(x='Nation', y='Value', data=data[data['Nation'].isin(top_nations)])
    plt.title(f'Value by Nationality (Nations {start_idx+1} to {end_idx})')
    plt.xlabel('Nationality')
    plt.ylabel('Value')
    plt.xticks(rotation=90)
    plt.show()

# Box plot of league vs. value
plt.figure(figsize=(14, 8))
sns.boxplot(x='League', y='Value', data=data)
plt.title('Value by League')
plt.xlabel('League')
plt.ylabel('Value')
plt.xticks(rotation=45)
plt.show()

club_values = data.groupby('Club')['Value'].sum().reset_index().sort_values(by='Value', ascending=False)

# Select the top 20 most valuable clubs
top_clubs = club_values.head(10)

plt.figure(figsize=(14, 8))
sns.barplot(x='Value', y='Club', data=top_clubs, palette='muted')
plt.title('Top 10 Most Valuable Football Clubs Overall')
plt.xlabel('Total Player Value (in million euros)')
plt.ylabel('Club')
plt.show()

<ipython-input-37-95d9564a5279>:7: FutureWarning:


Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `y` variable to `hue` and set `legend=False` for the same effect.

club_values = data.groupby('Club')['Value'].mean().reset_index().sort_values(by='Value', ascending=False).head(10)

plt.figure(figsize=(14, 8))
sns.barplot(x='Value', y='Club', data=club_values, palette='viridis')
plt.title('Top 10 Clubs by Average Player Value')
plt.xlabel('Average Value')
plt.ylabel('Club')
plt.show()

<ipython-input-38-ec62bed45d56>:4: FutureWarning:


Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `y` variable to `hue` and set `legend=False` for the same effect.

nation_counts = data['Nation'].value_counts()
valid_nations = nation_counts[nation_counts >= 25].index
data_filtered = data[data['Nation'].isin(valid_nations)]
nationality_values_filtered = data_filtered.groupby('Nation')['Value'].mean().reset_index().sort_values(by='Value', ascending=False)

plt.figure(figsize=(14, 8))
sns.barplot(x='Value', y='Nation', data=nationality_values_filtered, palette='viridis')
plt.title('Top Nations by Average Player Value (Minimum 25 Players)')
plt.xlabel('Average Value')
plt.ylabel('Nationality')
plt.show()

<ipython-input-39-9c4c840ca0ee>:7: FutureWarning:


Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `y` variable to `hue` and set `legend=False` for the same effect.

plt.figure(figsize=(14, 8))
sns.scatterplot(x='Age', y='Value', hue='Position', data=data)
plt.title('Market Value vs. Age')
plt.xlabel('Age')
plt.ylabel('Value')
plt.legend(title='Position', bbox_to_anchor=(1.05, 1), loc='upper left')
plt.show()

selected_columns = [col for col in data.columns if not (col.endswith('(17/18)') or col.endswith('(19/20)') or col.endswith('(18/19)'))]
for i in selected_columns:
  print(i)

Player
Club
Age
Position
Nation
Value
Contract Years Left
League
Squad (20/21)
MP (20/21)
Starts (20/21)
Min (20/21)
Total Mins/90 (20/21)
Gls (20/21)
Ast (20/21)
Non-Penalty Goals (20/21)
Penalties Scored (20/21)
Penalties Attempted (20/21)
Yellow Cards (20/21)
Red Cards (20/21)
Gls/90 (20/21)
Ast/90 (20/21)
(G+A)/90 (20/21)
Non-Penalty Goals/90 (20/21)
(Gls+Ast-Scored Penalties)/90 (20/21)
xG (20/21)
Non-Penalty xG (20/21)
xA (20/21)
Non-penalty xG+ xA (20/21)
xG/90 (20/21)
xA/90 (20/21)
(xG+xA)/90 (20/21)
Non-Penalty xG/90 (20/21)
(Non-Penalty xG+xA)/90 (20/21)
Total Shots (20/21)
Total Shots on Target (20/21)
Shots on Target% (20/21)
Shots/90 (20/21)
Shots on Target/90 (20/21)
Goals/Shots (20/21)
Goals/Shots on Target (20/21)
Avg Shot Distance (yds) (20/21)
Freekick Shots (20/21)
Non-Penalty xG/Shots (20/21)
Goals Scored minus xG (20/21)
Non-Penalty Goals Scored minus Non-Penalty xG (20/21)
Shot-Creating Actions (20/21)
Shot-Creating Actions/90 (20/21)
Passes Leading to Shot Attempt (20/21)
Set-Piece Leading to Shot Attempt (20/21)
Dribbles Leading to Shot Attempt (20/21)
Shots Leading to Shot Attempt (20/21)
Fouls Drawn Leading to Shot Attempt (20/21)
Defensive Actions Leading to Shot Attempt (20/21)
Goal Creating Actions (20/21)
Goal Creating Actions/90 (20/21)
Passes Leading to Goals (20/21)
Set-Piece Leading to Goals (20/21)
Dribbles Leading to Goals (20/21)
Shots Leading to Goals (20/21)
Fouls Drawn Leading to Goals (20/21)
Defensive Actions Leading to Goals (20/21)
Touches (20/21)
Touches in Defensive Penalty Box (20/21)
Touches in Defensive 3rd (20/21)
Touches in Midfield 3rd (20/21)
Touches in Attacking 3rd (20/21)
Touches in Attacking Penalty Box (20/21)
Touches in Open-play (20/21)
Total Successful Dribbles (20/21)
Total Attempted Dribbles (20/21)
Dribble Success % (20/21)
Total no. of Players Dribbles Past (20/21)
Total Nutmegs (20/21)
Total Carries (20/21)
Total Distance Carried the Ball (20/21)
Total Distance Carried the Ball in Forward Direction (20/21)
Total Carries in Forward Direction (20/21)
Carries into Final Third (20/21)
Carries into Attacking Penalty Box (20/21)
Total Failed Attempts at Controlling Ball (20/21)
Number of Times Tackled when Dribbling (20/21)
Number of Times Player was Pass Target (20/21)
Number of Times Received Pass (20/21)
% of Times Successfully Received Pass (20/21)
Progressive Passes Received (20/21)
Passes Completed (All pass-types) (20/21)
Passes Attempted (All pass-types) (20/21)
Pass Completion % (All pass-types) (20/21)
Total Distance of Completed Passes (All Pass-types) (20/21)
Total Distance of Completed Progressive Passes (All Pass-types) (20/21)
Passes Completed (Short Passes) (20/21)
Passes Attempted (Short Passes) (20/21)
Pass Completion % (Short Passes) (20/21)
Passes Completed (Medium Passes) (20/21)
Passes Attempted (Medium Passes) (20/21)
Pass Completion % (Medium Passes) (20/21)
Passes Completed (Long Passes) (20/21)
Passes Attempted (Long Passes) (20/21)
Pass Completion % (Long Passes) (20/21)
Total Assists (20/21)
xG Assisted (20/21)
Assist minus xG Assisted (20/21)
Completed passes that enter Final 3rd (20/21)
Completed passes that enter Penalty Box (20/21)
Completed Crosses that enter Penalty Box (20/21)
Total Completed Progressive Passes (20/21)
Total Number of Players Tackled (20/21)
Total Tackles Won (20/21)
Tackles in Defensive 3rd (20/21)
Tackles in Midfield 3rd (20/21)
Tackles in Attacking 3rd (20/21)
Number of Dribblers Tackled (20/21)
Times Dribbled Past + Total Tackles (20/21)
% of Dribblers Tackled (20/21)
Number of Times Dribbled Past (20/21)
Number of Pressing Actions (20/21)
Times Squad gained Possession within 5 seconds of Pressing Actions (20/21)
Successful Pressure % (20/21)
Number of Presses in Defensive Third (20/21)
Number of Presses in Midfield Third (20/21)
Number of Presses in Attacking Third (20/21)
Total Defensive Blocks (20/21)
Total Shots Blocked (20/21)
Goal Saving Blocks (20/21)
Times blocked a Pass (20/21)
Total Interceptions (20/21)
Total Players Tackled + Total Interceptions (20/21)
Total Clearances (20/21)
Mistakes leading to Opponent Shots (20/21)
2nd Yellow Cards (20/21)
Fouls Committed (20/21)
Fouls Drawn (20/21)
Offsides (20/21)
Crosses (20/21)
Interceptions (20/21)
Penalty Kicks Won (20/21)
Penalties Conceded (20/21)
Own Goals (20/21)
Total Loose Balls Recovered (20/21)
Aerial Duel Won (20/21)
Aerial Duel Lost (20/21)
% Aerial Duels Won (20/21)
G + A (20/21)
Outlier

passing_data = data[[
    'Age',
    'Position',
    'Value',
    'MP (20/21)',
    'Ast (20/21)',
    'Passes Completed (All pass-types) (20/21)',
    'Pass Completion % (All pass-types) (20/21)',
    'Passes Completed (Short Passes) (20/21)',
    'Pass Completion % (Short Passes) (20/21)',
    'Passes Completed (Medium Passes) (20/21)',
    'Pass Completion % (Medium Passes) (20/21)',
    'Passes Completed (Long Passes) (20/21)',
    'Pass Completion % (Long Passes) (20/21)',
    'Completed passes that enter Final 3rd (20/21)',
    'Completed passes that enter Penalty Box (20/21)',
    'Completed Crosses that enter Penalty Box (20/21)',
    'Total Assists (20/21)',
    'xG Assisted (20/21)',
    'Assist minus xG Assisted (20/21)',
    'Total Completed Progressive Passes (20/21)'
]]

defending_data = data[[
    'Age',
    'Position',
    'Value',
    'MP (20/21)',
    'Total Tackles Won (20/21)',
    'Tackles in Defensive 3rd (20/21)',
    'Tackles in Midfield 3rd (20/21)',
    '% of Dribblers Tackled (20/21)',
    'Number of Times Dribbled Past (20/21)',
    'Total Shots Blocked (20/21)',
    'Times blocked a Pass (20/21)',
    'Total Interceptions (20/21)',
    'Total Clearances (20/21)',
    'Mistakes leading to Opponent Shots (20/21)',
    'Own Goals (20/21)',
    'Total Loose Balls Recovered (20/21)',
    'Aerial Duel Won (20/21)',
    'Aerial Duel Lost (20/21)',
    '% Aerial Duels Won (20/21)',
    'Yellow Cards (20/21)',
    '2nd Yellow Cards (20/21)',
    'Red Cards (20/21)',
    'Number of Pressing Actions (20/21)',
    'Times Squad gained Possession within 5 seconds of Pressing Actions (20/21)',
    'Successful Pressure % (20/21)',
    'Total Defensive Blocks (20/21)',
    'Goal Saving Blocks (20/21)',
    'Total Players Tackled + Total Interceptions (20/21)'
]]


scoring_data = data[[
    'Age',
    'Position',
    'Value',
    'MP (20/21)',
    'Gls (20/21)',
    'Non-Penalty Goals (20/21)',
    'Penalties Scored (20/21)',
    'Penalties Attempted (20/21)',
    'Total Shots (20/21)',
    'Total Shots on Target (20/21)',
    'Goals/Shots (20/21)',
    'Goals/Shots on Target (20/21)',
    'Goal Creating Actions (20/21)',
    'Shot-Creating Actions (20/21)',
    'Total Attempted Dribbles (20/21)',
    'Dribble Success % (20/21)',
    'Number of Times Tackled when Dribbling (20/21)',
    'Number of Times Player was Pass Target (20/21)',
    '% of Times Successfully Received Pass (20/21)',
    'Progressive Passes Received (20/21)',
    'Goals Scored minus xG (20/21)',
    'Non-Penalty Goals Scored minus Non-Penalty xG (20/21)'
]]

passing_data_extended = data[['Age', 'Position', 'Value']]

seasons = ['20/21', '19/20', '18/19', '17/18']
for season in seasons:
    passing_data_extended = passing_data_extended.join(data[[f'MP ({season})',
                                                             f'Ast ({season})',
                                                             f'Passes Completed (All pass-types) ({season})',
                                                             f'Pass Completion % (All pass-types) ({season})',
                                                             f'Passes Completed (Short Passes) ({season})',
                                                             f'Pass Completion % (Short Passes) ({season})',
                                                             f'Passes Completed (Medium Passes) ({season})',
                                                             f'Pass Completion % (Medium Passes) ({season})',
                                                             f'Passes Completed (Long Passes) ({season})',
                                                             f'Pass Completion % (Long Passes) ({season})',
                                                             f'Completed passes that enter Final 3rd ({season})',
                                                             f'Completed passes that enter Penalty Box ({season})',
                                                             f'Completed Crosses that enter Penalty Box ({season})'
                                                            ]], how='left')

defending_data_extended = data[['Age', 'Position', 'Value']]
for season in seasons:
    defending_data_extended = defending_data_extended.join(data[[f'MP ({season})',
                                                                 f'Total Tackles Won ({season})',
                                                                 f'Tackles in Defensive 3rd ({season})',
                                                                 f'Tackles in Midfield 3rd ({season})',
                                                                 f'% of Dribblers Tackled ({season})',
                                                                 f'Number of Times Dribbled Past ({season})',
                                                                 f'Total Shots Blocked ({season})',
                                                                 f'Times blocked a Pass ({season})',
                                                                 f'Total Interceptions ({season})',
                                                                 f'Total Clearances ({season})',
                                                                 f'Mistakes leading to Opponent Shots ({season})',
                                                                 f'Own Goals ({season})',
                                                                 f'Total Loose Balls Recovered ({season})',
                                                                 f'Aerial Duel Won ({season})',
                                                                 f'Aerial Duel Lost ({season})',
                                                                 f'% Aerial Duels Won ({season})',
                                                                 f'Yellow Cards ({season})',
                                                                 f'2nd Yellow Cards ({season})',
                                                                 f'Red Cards ({season})',
                                                                 f'Number of Pressing Actions ({season})',
                                                                 f'Times Squad gained Possession within 5 seconds of Pressing Actions ({season})',
                                                                 f'Successful Pressure % ({season})',
                                                                 f'Total Defensive Blocks ({season})',
                                                                 f'Goal Saving Blocks ({season})',
                                                                 f'Total Players Tackled + Total Interceptions ({season})'
                                                                ]], how='left')

scoring_data_extended = data[['Age', 'Position', 'Value']]

seasons = ['20/21', '19/20', '18/19', '17/18']
for season in seasons:
    scoring_data_extended = scoring_data_extended.join(data[[f'MP ({season})',
                                                             f'Gls ({season})',
                                                             f'Non-Penalty Goals ({season})',
                                                             f'Penalties Scored ({season})',
                                                             f'Penalties Attempted ({season})',
                                                             f'Total Shots ({season})',
                                                             f'Total Shots on Target ({season})',
                                                             f'Goals/Shots ({season})',
                                                             f'Goals/Shots on Target ({season})',
                                                             f'Goal Creating Actions ({season})',
                                                             f'Shot-Creating Actions ({season})',
                                                             f'Total Attempted Dribbles ({season})',
                                                             f'Dribble Success % ({season})',
                                                             f'Number of Times Tackled when Dribbling ({season})',
                                                             f'Number of Times Player was Pass Target ({season})',
                                                             f'% of Times Successfully Received Pass ({season})',
                                                             f'Progressive Passes Received ({season})',
                                                             f'Goals Scored minus xG ({season})',
                                                             f'Non-Penalty Goals Scored minus Non-Penalty xG ({season})'
                                                            ]], how='left')

import seaborn as sns
import matplotlib.pyplot as plt

positions = ['midfield', 'Defender', 'attack']
for position in positions:
    aux_data = passing_data[passing_data['Position'] == position]
    aux_data.drop('Position', axis=1, inplace=True)
    aux_corr = aux_data.corr()
    plt.figure(figsize=(15, 8))
    plt.title(f'Correlation Matrix for {position} - Passing Data')
    sns.heatmap(aux_corr, annot=True, cmap='coolwarm', fmt=".2f")
    plt.show()

    aux_data = defending_data[defending_data['Position'] == position]
    aux_data.drop('Position', axis=1, inplace=True)
    aux_corr = aux_data.corr()
    plt.figure(figsize=(15, 8))
    plt.title(f'Correlation Matrix for {position} - Defending Data')
    sns.heatmap(aux_corr, annot=True, cmap='coolwarm', fmt=".2f")
    plt.show()

    aux_data = scoring_data[scoring_data['Position'] == position]
    aux_data.drop('Position', axis=1, inplace=True)
    aux_corr = aux_data.corr()
    plt.figure(figsize=(15, 8))
    plt.title(f'Correlation Matrix for {position} - Scoring Data')
    sns.heatmap(aux_corr, annot=True, cmap='coolwarm', fmt=".2f")
    plt.show()

<ipython-input-44-3ad88257a38a>:7: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

<ipython-input-44-3ad88257a38a>:15: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

<ipython-input-44-3ad88257a38a>:23: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

<ipython-input-44-3ad88257a38a>:7: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

<ipython-input-44-3ad88257a38a>:15: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

<ipython-input-44-3ad88257a38a>:23: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

<ipython-input-44-3ad88257a38a>:7: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

<ipython-input-44-3ad88257a38a>:15: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

<ipython-input-44-3ad88257a38a>:23: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

def plot_interactive_heatmap(data, position, title):
    corr_matrix = data.corr()
    fig = px.imshow(corr_matrix, text_auto=True, aspect="auto", title=title, color_continuous_scale='Viridis')
    fig.update_layout(width=1000, height=1000)  # Increase the size of the plot
    fig.show()

aux_data = passing_data_extended[passing_data_extended['Position'] == 'midfield']
aux_data = aux_data.drop('Position', axis=1)

corr_matrix = aux_data.corr()

plot_interactive_heatmap(aux_data, 'midfield', 'Interactive Correlation Matrix for Midfield - Passing Data')

aux_data = passing_data_extended[passing_data_extended['Position'] == 'Defender']
aux_data = aux_data.drop('Position', axis=1)

corr_matrix = aux_data.corr()

plot_interactive_heatmap(aux_data, 'Defender', 'Interactive Correlation Matrix for Defender - Passing Data')

aux_data = passing_data_extended[passing_data_extended['Position'] == 'attack']
aux_data = aux_data.drop('Position', axis=1)

corr_matrix = aux_data.corr()

plot_interactive_heatmap(aux_data, 'attack', 'Interactive Correlation Matrix for Attack - Passing Data')

aux_data = defending_data_extended[defending_data_extended['Position'] == 'midfield']
aux_data = aux_data.drop('Position', axis=1)

corr_matrix = aux_data.corr()

plot_interactive_heatmap(aux_data, 'midfield', 'Interactive Correlation Matrix for Midfield - Defending Data')

aux_data = defending_data_extended[defending_data_extended['Position'] == 'Defender']
aux_data = aux_data.drop('Position', axis=1)

corr_matrix = aux_data.corr()

plot_interactive_heatmap(aux_data, 'Defender', 'Interactive Correlation Matrix for Defender - Defending Data')

aux_data = defending_data_extended[defending_data_extended['Position'] == 'attack']
aux_data = aux_data.drop('Position', axis=1)

corr_matrix = aux_data.corr()

plot_interactive_heatmap(aux_data, 'attack', 'Interactive Correlation Matrix for Attack - Defending Data')

aux_data = scoring_data_extended[scoring_data_extended['Position'] == 'midfield']
aux_data = aux_data.drop('Position', axis=1)

corr_matrix = aux_data.corr()

plot_interactive_heatmap(aux_data, 'midfield', 'Interactive Correlation Matrix for Midfield - Scoring Data')

aux_data = scoring_data_extended[scoring_data_extended['Position'] == 'Defender']
aux_data = aux_data.drop('Position', axis=1)

corr_matrix = aux_data.corr()

plot_interactive_heatmap(aux_data, 'Defender', 'Interactive Correlation Matrix for Defender - Scoring Data')

aux_data = scoring_data_extended[scoring_data_extended['Position'] == 'attack']
aux_data = aux_data.drop('Position', axis=1)

corr_matrix = aux_data.corr()

plot_interactive_heatmap(aux_data, 'attack', 'Interactive Correlation Matrix for Attack - Scoring Data')

stat_groups = {'passing': passing_data, 'defending': defending_data, 'scoring': scoring_data}

for position in positions:
    print(f"Position: {position}")
    for group_name, group_data in stat_groups.items():
        print(f"Group: {group_name}")
        filtered_data = group_data[group_data['Position'] == position]
        filtered_data.drop('Position', axis=1, inplace=True)
        correlation_matrix = filtered_data.corr()
        top_5_features = correlation_matrix['Value'].abs().nlargest(11)
        top_5_features = top_5_features.drop('Value')
        plt.figure(figsize=(10, 10))
        sns.barplot(x=top_5_features.values, y=top_5_features.index)
        plt.title(f"Top 10 Features Correlated with 'Value' - {position} - {group_name}")
        plt.xlabel('Correlation (Absolute Value)')
        plt.ylabel('Feature')
        plt.show()

Position: midfield
Group: passing

<ipython-input-55-751b6f5954f4>:8: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

Group: defending

<ipython-input-55-751b6f5954f4>:8: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

Group: scoring

<ipython-input-55-751b6f5954f4>:8: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

Position: Defender
Group: passing

<ipython-input-55-751b6f5954f4>:8: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

Group: defending

<ipython-input-55-751b6f5954f4>:8: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

stat_groups = {'passing': passing_data_extended, 'defending': defending_data_extended, 'scoring': scoring_data_extended}

for position in positions:
    print(f"Position: {position}")
    for group_name, group_data in stat_groups.items():
        print(f"Group: {group_name}")
        filtered_data = group_data[group_data['Position'] == position]
        filtered_data.drop('Position', axis=1, inplace=True)
        correlation_matrix = filtered_data.corr()
        top_10_features = correlation_matrix['Value'].abs().nlargest(11)
        top_10_features = top_10_features.drop('Value')
        plt.figure(figsize=(10, 10))
        sns.barplot(x=top_10_features.values, y=top_10_features.index)
        plt.title(f"Top 10 Features Correlated with 'Value' - {position} - {group_name}")
        plt.xlabel('Correlation (Absolute Value)')
        plt.ylabel('Feature')
        plt.show()

Position: midfield
Group: passing

<ipython-input-56-5608b68c231f>:8: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

Group: defending

<ipython-input-56-5608b68c231f>:8: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

Group: scoring

<ipython-input-56-5608b68c231f>:8: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

Position: Defender
Group: passing

<ipython-input-56-5608b68c231f>:8: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

Group: defending

<ipython-input-56-5608b68c231f>:8: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

result = stats.f_oneway(data[data['League'] == 'Premier League']['Value'],
                        data[data['League'] == 'Serie A']['Value'],
                        data[data['League'] == 'La Liga']['Value'],
                        data[data['League'] == 'Bundesliga']['Value'],
                        data[data['League'] == 'Ligue 1']['Value'])

# Print ANOVA results
print("ANOVA results:")
print("F-statistic:", result.statistic)
print("p-value:", result.pvalue)

ANOVA results:
F-statistic: 29.889316060725868
p-value: 3.333603890933726e-24

data_filled_models = data_filled[data_filled['Position'] != 2]
data_filled_models = data_filled_models.drop(columns = ["Player", "Club"])
data_dropped_models = data_dropped[data_dropped['Position'] != 2]
data_dropped_models = data_dropped_models.drop(columns = ["Player", "Club"])

def get_top_correlated_features(df, target='Value', top_n=20):
    corr_data, mappings = encode_dataframe(df)
    corr_matrix = corr_data.corr()
    top_features = corr_matrix[target].abs().sort_values(ascending=False).head(top_n + 1).index.tolist()
    top_features.remove(target)
    return top_features

passing_top_features = get_top_correlated_features(passing_data_extended, 'Value')
defending_top_features = get_top_correlated_features(defending_data_extended, 'Value')
scoring_top_features = get_top_correlated_features(scoring_data_extended, 'Value')

all_top_features = list(set(passing_top_features + defending_top_features + scoring_top_features))
all_top_features.append('Value')

data_top_features_filled = data_filled_models[all_top_features].dropna()

X_filled = data_top_features_filled.drop('Value', axis=1)
y_filled = data_top_features_filled['Value']

data_top_features_dropped = data_dropped_models[all_top_features].dropna()

X_dropped = data_top_features_dropped.drop('Value', axis=1)
y_dropped = data_top_features_dropped['Value']

X_train, X_test, y_train, y_test = train_test_split(X_filled, y_filled, test_size=0.2, random_state=42)

print(f'Training set shape: X_train: {X_train.shape}, y_train: {y_train.shape}')
print(f'Testing set shape: X_test: {X_test.shape}, y_test: {y_test.shape}')

Training set shape: X_train: (1496, 58), y_train: (1496,)
Testing set shape: X_test: (374, 58), y_test: (374,)

linear_model = LinearRegression()
linear_model.fit(X_train, y_train)

# Make predictions on the test set
y_pred = linear_model.predict(X_test)

# Evaluate the model
mse = mean_squared_error(y_test, y_pred)
r2 = r2_score(y_test, y_pred)
print(f'Linear Regression - Mean Squared Error: {mse}')
print(f'Linear Regression - R-squared: {r2}')

Linear Regression - Mean Squared Error: 80968865505931.75
Linear Regression - R-squared: 0.5052717585324792

ridge_model = Ridge(alpha=1.0)
ridge_model.fit(X_train, y_train)

# Make predictions on the test set
y_pred = ridge_model.predict(X_test)

# Evaluate the model
mse = mean_squared_error(y_test, y_pred)
r2 = r2_score(y_test, y_pred)
print(f'Ridge Regression - Mean Squared Error: {mse}')
print(f'Ridge Regression - R-squared: {r2}')

Ridge Regression - Mean Squared Error: 80962356041762.8
Ridge Regression - R-squared: 0.5053115320397537

rf_model = RandomForestRegressor(n_estimators=100, random_state=42)
rf_model.fit(X_train, y_train)

# Make predictions on the test set
y_pred = rf_model.predict(X_test)

# Evaluate the model
mse = mean_squared_error(y_test, y_pred)
r2 = r2_score(y_test, y_pred)
print(f'Random Forest Regression - Mean Squared Error: {mse}')
print(f'Random Forest Regression - R-squared: {r2}')

Random Forest Regression - Mean Squared Error: 77731051509627.83
Random Forest Regression - R-squared: 0.5250551408805129

X_train, X_test, y_train, y_test = train_test_split(X_dropped, y_dropped, test_size=0.2, random_state=42)

print(f'Training set shape: X_train: {X_train.shape}, y_train: {y_train.shape}')
print(f'Testing set shape: X_test: {X_test.shape}, y_test: {y_test.shape}')

Training set shape: X_train: (492, 58), y_train: (492,)
Testing set shape: X_test: (124, 58), y_test: (124,)

linear_model = LinearRegression()
linear_model.fit(X_train, y_train)

# Make predictions on the test set
y_pred = linear_model.predict(X_test)

# Evaluate the model
mse = mean_squared_error(y_test, y_pred)
r2 = r2_score(y_test, y_pred)
print(f'Linear Regression - Mean Squared Error: {mse}')
print(f'Linear Regression - R-squared: {r2}')

Linear Regression - Mean Squared Error: 122622445494897.12
Linear Regression - R-squared: 0.604741972071713

ridge_model = Ridge(alpha=1.0)
ridge_model.fit(X_train, y_train)

# Make predictions on the test set
y_pred = ridge_model.predict(X_test)

# Evaluate the model
mse = mean_squared_error(y_test, y_pred)
r2 = r2_score(y_test, y_pred)
print(f'Ridge Regression - Mean Squared Error: {mse}')
print(f'Ridge Regression - R-squared: {r2}')

Ridge Regression - Mean Squared Error: 122513370554229.12
Ridge Regression - R-squared: 0.6050935614220219

rf_model = RandomForestRegressor(n_estimators=100, random_state=42)
rf_model.fit(X_train, y_train)

# Make predictions on the test set
y_pred = rf_model.predict(X_test)

# Evaluate the model
mse = mean_squared_error(y_test, y_pred)
r2 = r2_score(y_test, y_pred)
print(f'Random Forest Regression - Mean Squared Error: {mse}')
print(f'Random Forest Regression - R-squared: {r2}')

Random Forest Regression - Mean Squared Error: 120634165576451.61
Random Forest Regression - R-squared: 0.6111509422758417

	Player	Club	Age	Position	Nation	Value	Contract Years Left	League	Squad (20/21)	MP (20/21)	...	Offsides (17/18)	Crosses (17/18)	Interceptions (17/18)	Penalty Kicks Won (17/18)	Penalties Conceded (17/18)	Own Goals (17/18)	Total Loose Balls Recovered (17/18)	Aerial Duel Won (17/18)	Aerial Duel Lost (17/18)	% Aerial Duels Won (17/18)
0	Kylian Mbappe	Paris Saint-Germain	22	attack	France	144000000	1	Ligue 1	Paris S-G	31.0	...	23.0	62.0	1.0	1.0	0.0	0.0	98.0	1.0	4.0	20.0
1	Erling Haaland	Borussia Dortmund	21	attack	Norway	117000000	3	Bundesliga	Dortmund	28.0	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
2	Harry Kane	Tottenham Hotspur	28	attack	England	108000000	3	Premier League	Tottenham	35.0	...	43.0	24.0	7.0	1.0	0.0	0.0	124.0	69.0	111.0	38.3
3	Jadon Sancho	Manchester United	21	attack	England	90000000	5	Premier League	Dortmund	26.0	...	1.0	15.0	6.0	0.0	0.0	0.0	57.0	3.0	14.0	17.6
4	Mohamed Salah	Liverpool FC	29	attack	Egypt	90000000	2	Premier League	Liverpool	37.0	...	18.0	50.0	13.0	1.0	0.0	0.0	219.0	19.0	58.0	24.7

	Age	Value	MP (20/21)	Starts (20/21)	Min (20/21)	Total Mins/90 (20/21)	Gls (20/21)	Ast (20/21)	Non-Penalty Goals (20/21)	Penalties Scored (20/21)	...	Offsides (17/18)	Crosses (17/18)	Interceptions (17/18)	Penalty Kicks Won (17/18)	Penalties Conceded (17/18)	Own Goals (17/18)	Total Loose Balls Recovered (17/18)	Aerial Duel Won (17/18)	Aerial Duel Lost (17/18)	% Aerial Duels Won (17/18)
count	2075.000000	2.075000e+03	1910.000000	1910.000000	1910.000000	1910.000000	1910.000000	1910.000000	1910.000000	1910.000000	...	1154.000000	1156.000000	1156.000000	1156.000000	1156.000000	1154.000000	1156.000000	1156.000000	1156.000000	1052.000000
mean	26.712771	1.100229e+07	24.042932	18.586387	1658.567539	18.429104	2.383246	1.653403	2.124607	0.258639	...	4.283362	23.419550	18.346886	0.256055	0.225779	0.061525	171.566609	24.595156	24.645329	46.777391
std	4.242460	1.463476e+07	9.574457	10.485522	894.709390	9.941456	3.927010	2.296347	3.359727	0.999275	...	7.826631	29.142426	16.685891	0.632598	0.514779	0.286486	111.064881	30.633860	28.340609	18.826596
min	16.000000	9.000000e+04	1.000000	0.000000	1.000000	0.000000	0.000000	0.000000	0.000000	0.000000	...	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	1.000000	0.000000	0.000000	0.000000
25%	23.000000	2.250000e+06	17.000000	10.000000	914.000000	10.200000	0.000000	0.000000	0.000000	0.000000	...	0.000000	3.000000	5.000000	0.000000	0.000000	0.000000	84.000000	5.000000	7.000000	33.300000
50%	27.000000	5.400000e+06	26.000000	19.000000	1632.000000	18.150000	1.000000	1.000000	1.000000	0.000000	...	1.000000	11.000000	14.000000	0.000000	0.000000	0.000000	159.000000	15.000000	17.000000	48.550000
75%	30.000000	1.350000e+07	32.000000	28.000000	2405.000000	26.700000	3.000000	2.000000	3.000000	0.000000	...	4.750000	35.000000	27.000000	0.000000	0.000000	0.000000	242.000000	32.250000	32.000000	60.200000
max	43.000000	1.440000e+08	38.000000	38.000000	3420.000000	38.000000	41.000000	19.000000	33.000000	11.000000	...	56.000000	172.000000	93.000000	6.000000	3.000000	4.000000	555.000000	261.000000	285.000000	100.000000

	G + A (20/21)	Goal Creating Actions (20/21)
0	34.0	28.0
1	33.0	16.0
2	37.0	25.0
3	19.0	22.0
4	27.0	14.0
...	...	...
2070	NaN	NaN
2071	NaN	NaN
2072	NaN	NaN
2073	0.0	0.0
2074	NaN	NaN

	Player	Club	Age	Position	Nation	Value	Contract Years Left	League	MP (20/21)	Starts (20/21)	...	Times Squad gained Possession within 5 seconds of Pressing Actions (17/18)_94.0	Times Squad gained Possession within 5 seconds of Pressing Actions (17/18)_nan
0	Kylian Mbappe	Paris Saint-Germain	22	3	32	144000000	1	3	31.0	27.0	...	0	0
1	Erling Haaland	Borussia Dortmund	21	3	65	117000000	3	1	28.0	27.0	...	0	1
2	Harry Kane	Tottenham Hotspur	28	3	29	108000000	3	4	35.0	35.0	...	1	0
3	Jadon Sancho	Manchester United	21	3	29	90000000	5	4	26.0	24.0	...	0	0
4	Mohamed Salah	Liverpool FC	29	3	28	90000000	2	4	37.0	34.0	...	0	0

	Player	Club	Age	Position	Nation	Value	Contract Years Left	League	MP (20/21)	Starts (20/21)	...	Times Squad gained Possession within 5 seconds of Pressing Actions (17/18)_94.0
0	Kylian Mbappe	Paris Saint-Germain	22	3	32	144000000	1	3	31.0	27.0	...	0
2	Harry Kane	Tottenham Hotspur	28	3	29	108000000	3	4	35.0	35.0	...	1
3	Jadon Sancho	Manchester United	21	3	29	90000000	5	4	26.0	24.0	...	0
4	Mohamed Salah	Liverpool FC	29	3	28	90000000	2	4	37.0	34.0	...	0
5	Neymar	Paris Saint-Germain	29	3	11	90000000	4	3	18.0	15.0	...	0

Imports¶

Data Import and Some Preprocessing¶

Exploratory Data Analysis¶

Initial Correlation Plots¶

Correlation Plots for Value, Position and Important Statistics¶

Short Attempt at Predicting Player Values¶

DataFrame Filled with 0 Where Data Is Missing (hasn't played in that season, wasn't active, etc)¶

DataFrame Where We Drop Players Without a Full Record¶

Imports¶

Data Import and Some Preprocessing¶

Exploratory Data Analysis¶

Initial Correlation Plots¶

Plots Mostly Related to Player Position¶

Plots Mostly Related to Player Values¶

Correlation Plots for Value, Position and Important Statistics¶

Short Attempt at Predicting Player Values¶

DataFrame Filled with 0 Where Data Is Missing (hasn't played in that season, wasn't active, etc)¶

DataFrame Where We Drop Players Without a Full Record¶